Automatic merge from submit-queue
Add image pulling node e2e
Fixes#27007.
Based on #27309, will rebase after #27309 gets merged.
This PR added all tests mentioned in #27007:
* Pull an image from invalid registry;
* Pull an invalid image from gcr;
* Pull an image from gcr;
* Pull an image from docker hub;
* Pull an image needs auth with/without secrets.
For the imagePullSecrets test, I created a new gcloud project "authenticated-image-pulling", and the service account in the code only has "Storage Object Viewer" permission.
/cc @pwittrock @vishh
[]()
Automatic merge from submit-queue
Add description to created node images
Make it a little easier to see who to contact about important node e2e images.
Automatic merge from submit-queue
Updating federation up scripts to work in non e2e setup
Ref: https://github.com/kubernetes/kubernetes.github.io/pull/656
Updating the federation up scripts so that they work as per steps in https://github.com/kubernetes/kubernetes.github.io/pull/656.
Changes are:
* Updating the default namespace to be "federation" instead of "federation-e2e"
* Updated the kubeconfig context to be named "federation-cluster" instead of "federated-context"
* Fixing federation-up so that FEDERATION_IMAGE_TAG is set even when federation-up is run without running `e2e.go --up`. e2e-up.sh sets it here: 6a388d4a0d/hack/e2e-internal/e2e-up.sh (L44).
* Adding a "missingkey=zero" option to template parser. Without this, the parser adds `"<no value>"` at the place of an env var that is not set. With this change, it instead replaces it with the corresponding zero value (for ex "" for strings). This is required for the FEDERATION_DNS_PROVIDER_CONFIG env var.
cc @kubernetes/sig-cluster-federation @colhom @mml
Automatic merge from submit-queue
Implement first set of federated service e2e tests.
These tests are untested and there is no guarantee that they work. The ongoing auth problems is blocking these e2es from being tested and upon @quinton-hoole's request I am submitting them now.
Only the last commit here needs review.
Depends on #26953
cc @nikhiljindal @colhom @mfanjie @kubernetes/sig-cluster-federation
Automatic merge from submit-queue
Fix node e2e coreos kubelet cgroup detection
Fixes#26979#26431
The root issue, as best I can tell, is that cgroup detection does not work when the kubelet is started under an ssh session and the systemd `*Accounting` variables are set. I added additional logging and noted some differences in the cgroup slice names between those cadvisor returns and the kubelet detects for itself.
This difference does not occur if the kubelet is properly running under a unit. That environment is also a more common and sane environment.
See also discussion in #26903
cc @derekwaynecarr @vishh @pwittrock
Note that these tests are untested and there is no guarantee that they work.
The ongoing auth problems is blocking these e2es from being tested and upon
@quinton-hoole's request I am submitting them now.
Automatic merge from submit-queue
Add pending pod check in cluster autoscaler e2e tests
The tests should wait until all pods are running before declaring a success and resizing the mig.
cc: @fgrzadkowski @piosz @jszczepkowski
Automatic merge from submit-queue
volume integration: wait for PVs before creating PVCs
The test should wait until all volumes are processed by volume controller (i.e. in the controller cache) before creating a PVC.
Without that, the "best" matching PV could not be in the cache and controller might bind the PVC to suboptiomal one.
This fixes integration test flake "Bind mismatch! Expected pvc-2 capacity 50000000000 but got pvc-2 capacity 52000000000".
Fixes#27179 (together with #26894)
Automatic merge from submit-queue
Fix integration pv flakes
There are two fixes in this PR:
- run tests in separarate functions and use objects with different names, otherwise events from the beginning of the function are caught later when we watch for events of a different PV/PVC
- don't set PV.Spec.ClaimRef.UID of pre-bound PVs. PVs with UID set are considered as bound and they are deleted/recycled when appropriate PVC does not exists yet.
Fixes#26730 and probably also ~~#26894~~ #26256
The test should wait until all volumes are processed by volume controller (i.e.
in the controller cache) before creating a PVC.
Without that, the "best" matching PV could not be in the cache and controller
might bind the PVC to suboptiomal one.
This fixes integration test flake "Bind mismatch! Expected pvc-2 capacity
50000000000 but got pvc-2 capacity 52000000000".
Automatic merge from submit-queue
e2e: actually check error when we fail to GET the scheduled pod
and GET the pod before we try to gracefully delete it.
and add a debug log.
ref #26224
Automatic merge from submit-queue
Considering all nodes for the scheduler cache to allow lookups
Fixes the actual issue that led me to create https://github.com/kubernetes/kubernetes/issues/22554
Currently the nodes in the cache provided to the predicates excludes the unschedulable nodes using field level filtering for the watch results. This results in the above issue as the `ServiceAffinity` predicate uses the cached node list to look up the node metadata for a peer pod (another pod belonging to the same service). Since this peer pod could be currently hosted on a node that is currently unschedulable, the lookup could potentially fail, resulting in the pod failing to be scheduled.
As part of the fix, we are now including all nodes in the watch results and excluding the unschedulable nodes using `NodeCondition`
@derekwaynecarr PTAL
Automatic merge from submit-queue
Port the downward api test to the node e2e suite
Also extend the framework to allow a custom client config loading function, so
that the node e2e suite can reuse the same framework across tests.
This fixes#26609
/cc @timstclair @pwittrock
Automatic merge from submit-queue
Add e2e test for node problem detector
Based on https://github.com/kubernetes/node-problem-detector/pull/12.
This PR added e2e test for node problem detector. It starts a node problem detector with test configuration, and test its functionality.
Question:
* Should this be a node e2e test or e2e test?
* Should we mark this as Serial? The test should not affect other tests, but it will add a `TestConditon` in node conditions and remove in cleanup.
@dchen1107
/cc @kubernetes/sig-node
[]()
Automatic merge from submit-queue
Listing pods only once when getting pods for RS in deployment
Fixes#26834
1. Avoid ranging over RSes and then `List` pods of each RS. Instead, `List` pods of the deployment once, and then filter pods of each RS.
2. Avoid using clientset to `List` pods in deployment controller. Use podStore instead. (TODO in some functions because the unit tests don't have podStore.)
@kubernetes/deployment
[]()
Automatic merge from submit-queue
support for mounting local-ssds on GCI
This change adds support for mounting local ssds on GCI.
It updates the previous container-vm behavior as well to
match that for GCI nodes by mounting the local-ssds under
the same path (/mnt/disks/ssdN).
@vulpecula @roberthbailey @andyzheng0831 @kubernetes/goog-image
This reverts commit 2494c77972.
The previous commit, which launches the kubelet under `systemd-run`,
fixes an error in detecting the kubelet's cgroup stats.
Fixes#26979
Automatic merge from submit-queue
Fix GKE upgrade e2e util.
containers command group at HEAD no longer accepts --zone. Flag
has to be specified after subcommand group. Fix#27011
This change adds support for mounting local ssds on GCI.
It updates the previous container-vm behavior as well to
match that for GCI nodes by mounting the local-ssds under
the same path (/mnt/disks/ssdN).
Automatic merge from submit-queue
Enable WatchCache in test/integration/ tests
We already run cmd/integration/ with watch cache on. We should also run tests in test/integration/ with watch cache on.
@wojtek-t @lavalamp
Automatic merge from submit-queue
Add a custom main instead of the standard test main, to reduce stack …
Adds a custom test main handler (see: `TestMain` in https://golang.org/pkg/testing/ for details)
Partial fix for https://github.com/kubernetes/kubernetes/issues/25965
This does the standard timeout, but strips non-kubernetes stacks out of the stack trace (e.g. it filters things like:
```
goroutine 466 [IO wait, 7 minutes]:
net.runtime_pollWait(0x7fd74c4672c0, 0x72, 0xc821614000)
/usr/local/go/src/runtime/netpoll.go:160 +0x60
net.(*pollDesc).Wait(0xc8215c21b0, 0x72, 0x0, 0x0)
/usr/local/go/src/net/fd_poll_runtime.go:73 +0x3a
net.(*pollDesc).WaitRead(0xc8215c21b0, 0x0, 0x0)
/usr/local/go/src/net/fd_poll_runtime.go:78 +0x36
net.(*netFD).Read(0xc8215c2150, 0xc821614000, 0x1000, 0x1000, 0x0, 0x7fd74c491050, 0xc820014058)
/usr/local/go/src/net/fd_unix.go:250 +0x23a
net.(*conn).Read(0xc820a5a090, 0xc821614000, 0x1000, 0x1000, 0x0, 0x0, 0x0)
/usr/local/go/src/net/net.go:172 +0xe4
net/http.noteEOFReader.Read(0x7fd74c465258, 0xc820a5a090, 0xc8215f0068, 0xc821614000, 0x1000, 0x1000, 0x405773, 0x0, 0x0)
/usr/local/go/src/net/http/transport.go:1687 +0x67
net/http.(*noteEOFReader).Read(0xc8215ae1a0, 0xc821614000, 0x1000, 0x1000, 0xc82159ad1d, 0x0, 0x0)
<autogenerated>:284 +0xd0
bufio.(*Reader).fill(0xc8202a2b40)
/usr/local/go/src/bufio/bufio.go:97 +0x1e9
bufio.(*Reader).Peek(0xc8202a2b40, 0x1, 0x0, 0x0, 0x0, 0x0, 0x0)
/usr/local/go/src/bufio/bufio.go:132 +0xcc
net/http.(*persistConn).readLoop(0xc8215f0000)
/usr/local/go/src/net/http/transport.go:1073 +0x177
created by net/http.(*Transport).dialConn
/usr/local/go/src/net/http/transport.go:857 +0x10a6
```
We may want to get even more aggressive in the future.
@kubernetes/sig-testing
Automatic merge from submit-queue
Move quota usage testing for loadbalancers into unit tests
Fixes https://github.com/kubernetes/kubernetes/issues/26319
* moved testing for node port and load balancer usage in quota to unit tests
* remove node port and node port -> loadbalancer service testing out of e2e
* covered already in replenishment_controller_test scenario
Given the time it takes to even allocate a load balancer, it seems better to test that outside of this test case to avoid unnecessary flakes.
/cc @bprashanth
Automatic merge from submit-queue
Flake 26210: decouple explicit access from port 80
Flake #26210 only happens for port 80. To decouple the possible causes, all
tests with explicit port 80 are moved to port 1080 (these were 80% of the flakes).
The urls without a specified port (which map to port 80 though) are left untouched.
If port 1080 does not show up as flake now, there is really a connection to the
actual port number.
Automatic merge from submit-queue
Reduce huge amount of logs in large cluster tests
When running tests in 2000-node clusters, I got more than 100.000 lines like this:
```
Jun 8 01:03:11.850: INFO: Condition NetworkUnavailable of node gke-gke-large-cluster-default-pool-1-03ee5a12-knrw is true instead of false. Reason: NoRouteCreated, message: Node created w ithout a route
```
that doesn't give much value.
This is PR is reducing the number of logs.
Also includes other improvements:
- Makefile rule to run tests against remote instance using existing host or image
- Makefile will reuse an instance created from an image if it was not torn down
- Runner starts gce instances in parallel with building source
- Runner uses instance ip instead of hostname so that it doesn't need to resolve
- Runner supports cleaning up files and processes on an instance without stopping / deleting it
- Runner runs tests using `ginkgo` binary to support running tests in parallel
Now that GCE routes take an extremely long time to come up and there's
a variance in "Ready" and "Schedulable", start cherry-picking tests
where we really want to have all nodes routable/schedulable for
testing. Adding logging. This will increase test times on large
clusters but should have 0 impact on normal testing.
Flake #26210 only happens for port 80. To decouple the possible causes, all
tests with explicit port 80 are moved to port 1080 (these were 80% of the flakes).
The urls without a specified port (which map to port 80 though) are left untouched.
If port 1080 does not show up as flake now, there is really a connection to the
actual port number.
Automatic merge from submit-queue
Mark runtime conformance tests as flaky and not run them in jenkins CI.
For #26809
@pwittrock As discussed offline, marking runtime tests as flaky for now. I'm not sure if those tests are required. Testing docker in every Kubernetes PR is un-necessary.
These tests can be run periodically in a separate CI. AFAIK, these tests don't seem to exercise any kube features.
When we create a PV, we should created it withoud Spec.ClaimRef.UID.
In rare cases, when 'PV added' event with UID is processed before 'PVC
added' (created by for loop few lines above), the controller does not know
a PVC with this UID and considers the PV as released. Reclaim policy is
then executed and the PV is deleted and it's never bound.
With UID="", the controller waits for the PVC to get created and binds
it.
Different tests should use different objects and watchers - I noticed
sometimes an event from old tests leaked into subsequent test in the
same function.
And add some logs.
Automatic merge from submit-queue
Update Node e2e Core OS image to run systemd with CPU & Memory accounting enabled by default
cc @derekwaynecarr
For #26289
Automatic merge from submit-queue
volume controller: add configurable integration test to stress the binder
The test tries to bind configured nr. of PVs to the same nr. of PVCs. '100' is used by default, which should take ~1-3 seconds (depends on log level). Periodic sync is needed in rare cases, which may add another 10 seconds. - cache from #25881 will help here and sync should not be needed at all.
The test is configurable and may be reused to measure binder performance. Set KUBE_INTEGRATION_PERSISTENTVOLUME_* env. variables as described in persistent_volume_test.go and run the tests:
```
# compile
$ cd test/integration
$ godep go test -tags 'integration no-docker' -c
# run the tests
$ KUBE_INTEGRATION_PERSISTENTVOLUME_SYNC_PERIOD=10s KUBE_INTEGRATION_PERSISTENTVOLUME_OBJECTS=1000 time ./integration.test -test.run TestPersistentVolumeMultiPVsPVCs -v 2
```
Log level '2' is useful to get timestamps of various events like 'TestPersistentVolumeMultiPVsPVCs: start' and 'TestPersistentVolumeMultiPVsPVCs: claims are bound'.
Automatic merge from submit-queue
kubelet e2e: enforce that image prepulling must finish before the test
The image prepulling pod calls docker directly to pull images. If the pod
hasn't finished before running the resource usage tracking test, there'd be a
cpu spike in docker. We'd rather wait and fail if this is the case, before
running the test.
The image prepulling pod calls docker directly to pull images. If the pod
hasn't finished before running the resource usage tracking test, there'd be a
cpu spike in docker. We'd rather wait and fail if this is the case, before
running the test.
